Dual-Role Global Markets Simulation

Analytical Question

How do inventory control, execution choice, client flow and dynamic exposure management translate into absolute and risk-adjusted outcomes across repeated buy-side and sell-side market simulations?

Evidence Design

The project uses timed AmplifyME Finance Accelerator – Global Markets sessions completed on 15, 22, 27 and 29 July and 3 August 2026. Each session had a different cohort, market path and news sequence. Comparisons are therefore descriptive rather than controlled, and platform scores are reported separately from financial outcomes.

The strongest role-specific results occurred on different dates. The principal buy-side evidence comes from 29 July; the principal sell-side evidence comes from 27 July. The 3 August event is retained because it provides the strongest same-event evidence across both mandates, not because it supersedes either flagship result.

Principal Buy-Side Result — 29 July

The asset-management session scored 93.05% and appeared 2nd of 102 on the displayed leaderboard. Event administration subsequently voided the displayed first-place result, leaving this as the highest valid score. The displayed rank and post-event status are both retained; the result is not described as an unqualified first place.

The portal reported $3,801,651 of P&L, a 19.0% return and a 1.20 Sharpe ratio. Its 100% ROI field is a proprietary subscore, not a 100% investment return. The detailed and leaderboard screens also use different labels for the $3.80m field, and the displayed portfolio value does not mechanically reconcile with P&L and commission.

Allocation and Risk Interpretation

Average allocation was 70.2% equities and 29.8% cash; the portfolio finished 99.5% in equities. That exposure captured the main return-generating phase, but it also left the result dependent on a highly concentrated final allocation. No transaction timestamps or security-level weights are reconstructed from the portal chart.

Principal Sell-Side Result — 27 July

The dealer-trading session ranked 8th of 93 with an 88.89% composite score. It produced $842,052 of trading P&L and $153,635 of commission, for $995,687 of portal-reported total simulated P&L. No exchange or client fat-finger incident was recorded.

Average net exposure was -$10,410,039 and average contract exposure was -37,152. Stronger financial and platform outcomes therefore do not imply that dealer inventory risk fell on every measure.

Same-Event Cross-Role Result — 3 August

The asset-management phase ranked 5th of 93 with an 89.68% score and a $3,330,183 platform trading-P&L field. The sell-side phase ranked 9th of 96 with an 89.04% score, $893,072 of trading P&L and $202,035 of commission.

The sell-side total of $1,095,107 is derived by adding trading P&L and commission because the portal did not show a total field. The significance of the event is simultaneous top-decile performance across both roles; it is not presented as another flagship.

Session Record and Failure Evidence

The ledger preserves both stronger and weaker sessions. In particular, the 29 July sell-side phase generated $192,145 of commission but a $496,309 trading loss, producing total simulated P&L of -$304,164. This is useful evidence that client-process and operational metrics can remain credible while dealer inventory produces a poor financial result.

Measured Lessons

• Buy-side return, Sharpe ratio, execution score and composite score all improved between the 22 and 29 July principal sessions, but final concentration also increased materially.

• Sell-side commission, trading P&L and composite score improved in the principal 27 July session, while average dealer exposure remained substantial.

• A formal late-session de-risking rule would make the buy-side process less dependent on remaining almost fully invested into the close.

• Client facilitation, commission generation and speculative dealer inventory should be monitored as separate sources of outcome.

• Proprietary platform subscores cannot be treated as causes of ranking or financial performance without the weighting formula.

Limitations

Related Work

Source: participant-level AmplifyME portal results, displayed leaderboards and supplied event screenshots. No full transaction export or complete scoring formula was available. All P&L figures refer to simulated rather than real capital.